Library

Neural relations: your Library, everywhere you work

Neural relations surfaces saved items related to whatever you're writing, organizing, or discussing — look for the brain icon in notes, boards, and chat.

·4 min read

Neural relations is your Library resurfacing itself while you work. Wherever you see the brain icon — in a note, on a board, in a chat — Eden is quietly matching what you're doing against everything you've saved and offering the items that relate.

Most saved things are never seen again, because remembering to search is the hard part. Neural relations removes that step. You don't ask; Eden watches what you're working on and brings the relevant saves to you.

It runs on the same AI search that powers the rest of your Library, so it matches by meaning, not keywords. Write about focus and it finds the highlight about deep work, even though the word "focus" never appears in it.

Where it appears

The brain icon shows up in three places, and it does the same job in each: suggest related items, one click to use them.

  • In a note. While you write, the Neural panel sits beside your document and updates as your paragraph evolves. Click a suggestion to open it in a side pane, or hit the @ action to weave it into your text as a mention — which also connects the two items.
  • On a board. The brain icon in the board's bottom toolbar suggests Library items related to the board's title, sections, and what's already on it. One click adds an item to the board.
  • In a chat. The brain icon in the chat header suggests items related to the current conversation. One click attaches an item as context, so Eden's next reply draws on your own material. More in chatting with your Library.

How suggestions are chosen

Every suggestion comes from your own Library — your links, notes, PDFs, saved posts, and highlights. Eden reads the context of the surface you're on (the paragraph you're writing, the board you're building, the exchange you just had) and runs a meaning-based search across your saves.

A few things it deliberately does:

  • It never suggests what's already there. Items already on the board, already woven into the note, or already attached to the chat stay out of the list.
  • It updates as you go. Move to a new paragraph or a new question and the suggestions follow. There's a short settle delay so the panel doesn't flicker while you type.
  • It suggests, never acts. Nothing is added, woven, or attached until you click. The panel is a shelf, not an autopilot.

Choosing what gets suggested

You decide which kinds of items the panel may offer. Click the gear in the panel's header and untick any type — documents, highlights, links, social posts, PDFs, and so on. Prefer only external sources? Untick Documents and Cards, and the panel sticks to things like highlights, articles, and saved posts.

The choice is one setting shared by every surface — notes, boards, and chat all follow it — and it's also available under Settings → Preferences → Neural relations. Everything is shown by default.

Why it compounds

Neural relations is the payoff for the save-first habit. Every article you clip, post you save, and highlight you make becomes something Eden can hand back at exactly the moment it's useful — while writing the draft, filling the board, or thinking through a question in chat. The more you save, the better every surface gets.

Where to go next

Still stuck?

Email us. A real person reads every message.

Tell us what you tried and where you got stuck. We answer within one business day.

[email protected]